(2) The Behavior of a2-Plasmin Inhibitor in Fibrinolytic States NOBuo AoKi, MASAAKI MOROI, MICHIO MATSUDA, and KAZUKO TACHIYA From the Department of Medicine and the Institute of Hematology, Jichi Medical School, Minamikawachi-Machi, Tochigi-ken 329-04, Japan A B S T R A C T Human plasma a2-plasmin inhibitor in fibrinolytic states was studied using immunochemical methods and radioiodinated plasminogen. The concentration and activity of plasma a2-plasmin inhibitor decreased when urokinase was added to plasma in vitro or infused intravenously in man. The decrease was associated with the appearance of plasmin-a2-plasmin inhibitor complex which subsequently disappeared from the circulation in a short time. A decrease of other major inhibitors, such as a2-macroglobulin and a1-antitrypsin, was not observed when the amount of urokinase added or infused was relatively small, and conversion of plasminogen to plasmin was not extensive. The formation of plasmina2-macroglobulin complex was observed only when plasma plasminogen was activated with a larger amount of urokinase, and after most of the a2-plasmin inhibitor was consumed by forming complexes with plasmin. The formation of plasmin-al-antitrypsin complex was not observed even in the highly activated plasma unless exogenous plasmin was added to the plasma. a2-Plasmin inhibitor was the only inhibitor of which the concentration in plasma was significantly decreased in patients with disseminated intravascular coagulation and fibrinolysis among the major plasmin inhibitors in plasma. The most reactive inhibitor for regulating plasma fibrinolysis very likely is a2-plasmin inhibitor. INTRODUCTION A plasma fraction is known which inhibits plasminogen activator-induced clot lysis efficiently but posPart of this data was presented at the Intemational Society of Hematology Meeting, 8 September 1976. The research was carried out according to the provisions of the Declaration of Helsinki, and informed consent was obtained from each one of the patients. Receivedfor publication 23 September 1976 and in revised form 1 April 1977.. sesses only a small part of the whole capacity of plasma to neutralize plasmin activity (1, 2). This fraction is called antiactivator, since it was found to be different from the major known plasmin inhibitors such as a2-macroglobulin (a2-M)l and a1-antitrypsin (a,-AT), both which strongly inhibit plasmin on incubation but exert little effect on activator-induced clot lysis (1, 3). This antiactivator protein was purified, and its properties were investigated (4). The antiactivator was ultimately found to be primarily a plasmin inhibitor belonging to a2-globulin, and the term. a2-plasmin inhibitor (a2-PI) or a2-proteinase inhibitor is considered to be appropriate. a2-PI was shown to be different from a2-M, CI inactivator, inter-a-trypsin inhibitor, a1-antichymotrypsin, a,-AT, and antithrombin III (4). a2-PI inhibits plasmin almost instantaneously, and the inhibition of plasminogen activation by the inhibitor was found to be mainly due to the very rapid inactivation of plasmin formed (4). In this paper, we consider that this unique biological property of a2-PI has an important role in regulating fibrinolysis in vivo. METHODS Plasma. Human blood was collected from antecubital veins into a 0.1-vol of 3.8% trisodium citrate and was centrifuged at 2,000 g at tip for 20 min to prepare plateletpoor plasma. Preparation of purified plasminogen and plasmin. Plasminogen was prepared from human fresh plasma according to the method of Brockway and Castellino (5). Plasminogen fraction 2, which was eluted after fraction 1, was used for all the studies. Plasmin was prepared by activating plasminogen with urokinase-coupled Sepharose (Pharmacia Fine Chemicals, Div. of Pharmacia, Inc., Piscataway, N.J.) according to the method previously described (4). Activity of plasminogen and plasmin was assayed according to the caseinolytic method of Robbins and Summaria (6). Protein concentra'Abbreviations used in this paper: a2-M, a2-macroglobulin; a2-PI, a2-plasmin inhibitor; a1-AT, a,-antitrypsin; DIC, disseminated intravascular coagulation.. The Journal of Clinical Investigation Volume 60 August 1977 -361 -369. 361.
(3) tion of purified plasminogen was determined by measuring U/ml per A280 (4), in a series of preparations and appeared as the absorbance of the plasminogen solution (pH 7.4) at 280 a single band on sodium dodecyl sulfate polyacrylamide nm and by converting absorbance to protein concentration gel electrophoresis. The protein concentrations of the purified inhibitor were determined by measuring the absorbance using E19m = 17.0 for purified plasminogen (7). Urokinase preparations. Urokinase preparations used for of the inhibitor solution (pH 7.4) at 280 nm and converting intravenous infusion were obtained from Green Cross Corp., absorbance to protein concentration using E""m = 7.03 for Osaka and from Mochida Pharmaceutical Co., Tokyo. Ac- pure inhibitor (4). The extinction coefficient, EF^3m, was detivity was expressed by intemational units based on the rived from the measurement of absorbance of the solution of standard preparations supplied by the World Health Organ- known weight of the purified inhibitor, which had been ization. The unit is nearly equivalent to the Committee of extensively dialyzed to remove salts, lyophilized, and then dried in a vacuum at 1 10°C over P205 for 4 days. Thrombolytic Agents unit. The concentrations of a2-M and a,-AT in human plasma or Assay for plasma plasminogen and fibrinogen. Plasminby the single radial immunodifogen in plasma was measured after destruction of anti- serum were determined plates (Behring-Werke AG). partigen using technique fusion be assayed to All plasmas (8). acidification plasmin by was performed electrophoresis Antigen-antibody-crossed were first acidified by the addition of the equal volume of 0.166 N HCI, allowed to stand for 15 min, reneutralized essentially according to Laurell (13). 30 ul of antigen x 10 mm) in with 0.166 N NaOH, and then subjected to the caseinolytic solution was placed in a transverse slit (1.5buffer, pH 8.6, barbital in agarose layer (0.8% agarose the and in Remmert method (6). The results were expressed on a glass slide Cohen casein units (9). The fibrinogen content of plasma ionic strength 0.05, and 1.8 mm thickness) (2.5 x 7.5 cm). The electrophoresis was performed for about was assayed by the method of Ratnoff and Menzie (10). h with a constant current of 2 mA/cm width. After2Y2 human against antisera Immunochemical methods. Rabbit the gel slip was transferred to a second glass plate a,-AT and a2-M were obtained from Behring-Werke AG, Mar- wards (5 x 7.5 cm). The remainder of the second plate was then burg/Lahn, West Germany. Rabbit antiserum against human covered with a layer of agarose containing 1% antiserum a2-PI was prepared by immunizing rabbits with a purified against a2-PI, 0.4% antiserum against plasminogen, or 2% preparation of the inhibitor (4). About 1 mg of protein in 1 ml against a,-AT, separately. The second electrophoreof 0.05 M Tris-HCl, pH 7.4, containing 0.15 M NaCl antiserum by applying the electric current (1 mA/ was mixed completely with an equal volume of Freund's sis was performed of 900 to the direction of the first angle an at width) cm divided in equally complete adjuvant and was injected separation. The time of the second run was volumes intracutaneously into each footpad, shoulder, and electrophoretic 10h. The buffer used for preparation of agarose gel and thigh of a rabbit. 3 wk later, the injection was repeated electrophoretic was barbital buffer (pH 8.6, ionic strength with 0.5 mg protein. Immune sera were harvested 2-5 wk 0.05). The platesrunwere washed with saline to remove nonafter the second injection, and were absorbed with an a2-PIprotein and dried. The immune precipitates were free plasma fraction in step 4 of the purification procedures precipitated with amido black. of the inhibitor described previously (4). The fraction had then stained activity. Antiplasmin activity of antiplasmin 'a2M' Sepharose been passed through the plasminogen-coupled the modified method of the fibrinoby estimated was plasma column repeatedly to remove a2-PI completely. 1 vol of this lytic assay described previously 0.1 ml of 30 ,ug/ml (0.52 plasma fraction, absorbancy of which was 17 at 280 nm, was casein U/ml) plasmin in 25% (1).glycerol mixed and added to 40 vol of antiserum, incubated at 37°C for 15 incubated at 37°C for 30 min with 0.1 ml ofwas test plasma the min, and stored at 4°C for 48 h. The resulting precipitate which had been diluted 20-fold with barbital-buffered was removed by centrifugation and discarded. The supernate (5.2 mM barbital 0.14 M NaCl, pH 7.4). The mixturesaline was stored in a small aliquot with 0.1% of sodium azide as then transferred to an ice-water bath, diluted with 0.6 mlwas of a preservative. Antiserum thus obtained was shown to be cold barbital-buffered saline, and left for 3 min. Subsemonospecific by double immunodiffusion and immuno- quently, 0.1 ml of fibrinogen solution followed by 0.1 ml of electrophoresis as previously described (4). The IgG fraction of antisera was obtained by DEAE-cellulose chromatography (11). Rabbit antiserum against human plasminogen was prepared by immunizing rabbits with a purified preparation of plasminogen. Immunization procedure was essentially the same as the one used for antiserum against a2-PI described above. About 10 mg of protein was used for the first injection LU and 1.5 mg for the second injection. Immune sera were abc sorbed with plasminogen-free human plasma which had t: passed through a column of lysine-Sepharose (4). 1 vol of u fivefold-diluted plasminogen-free plasma was added to 9 vol cyof antiserum, and the precipitate formed was removed. 0 The concentration of a2-PI in human plasma or serum was x determined by the one-dimensional method of Laurell (12) (D z using 1% antiserum in agarose (0.8% in barbital buffer, pH 8.6, ionic strength 0.05). 10 p1 of antigen solution (purified inhibitor, plasma, or serum) was applied to the electro12 8 6 4 2 IO phoresis which was run with a constant current of 1 mA/cm width for 7 h. Dilutions of known amounts of the purified OF INHIBITOR NTRATION (mg/lOOml) CONCE inhibitor were made with Tris-buffered saline (Tris 0.05 M, NaCl 0.15 M, pH 7.4) containing 3% wt/vol bovine serum FIGURE I A reference curve for measuring the concebtration albumin and used for calibration. There was a linear rela- of a2-PI by the one-dimensional method of Laurell (12). A tionship between peak heights and amounts of antigen ap- linear relationship between concentration of purified aL2-PI plied (Fig. 1). The purified inhibitor used for calibration and length of precipitate (peak heights) was obtained. For was the one which had the highest specific activity, 1,933 details see the text. LU CL. LL. LLI. 362. N. Aoki, M. Moroi, M. Matsuda, and K. Tachiya.
(4) 20 U/ml bovine thrombin (Parke, Davis & Company, Detroit, Mich.) was added, and the solution was transferred to a 370C water bath, and the clot lysis time was recorded. Fibrinogen solution used was 2% Cohn's fraction I of human plasma (Green Cross Corp., Osaka) which had been treated with lysine-Sepharose to remove plasminogen (14) and contained 1.6 g/100 ml of fibrinogen. The control was run by replacing the test plasma with buffered saline. The difference between the control measurement and the test run, expressed in fibrinolytic plasmin units as defined previously (1), divided by the volume (0.1 ml) of the test sample, multiplied by the dilution factor 20 represents the antiplasmin units per milliliter. The determinations of the activity of plasma samples from 12 healthy adults revealed that the mean ±2 SD of normal values are 83±42 U/ml. The antiplasmin activity measured by this method is referred to as a2-M antiplasmin activity, since the method measures mainly the antiplasmin activity of a2-M and is insensitive to the change of a2-PI activity (1-3). Insensitivity of the method to the change of a2-PI may be explained by the fact that the method measures nearly total capacity of antiplasmin activity of whole plasma (1), and a2-PI contributes only 2.5% on average of the total antiplasmin activity (see Discussion). Assay of ag-PI activity. Activity of a2-PI in plasma or serum was assayed by the method based on the inhibitory activity of plasma or serum to activator-induced clot lysis (1, 2). Activator-induced clot lysis is principally inhibited by a2-PI but also by a2-M to a small degree (2). When plasminogen-free plasma prepared by lysine-Sepharose (4) was chromatographed on Sephadex G-200 (Pharmacia Fine Chemicals, Div. of Pharmacia, Inc.) as described previously (2), and each fraction of the effluent was analyzed for inhibitory activity to urokinase-induced clot lysis (2), the a2-M fraction which emerged in the first protein peak accounted for approximately 10-20% of the total activity of plasma depending on a2-M concentration in each plasma sample used. The inhibitory activity exerted by the a2-M fraction was diminished to approximately half by freezing and thawing the fraction, whereas inhibitory activity of a2-PI fraction which emerged later in the chromatography was not affected by a freezing-thawing procedure. This was also shown by the study using purified a2-M (15) and purified a2-PI. Inhibitor activity of 8 mg/ml a2-M was 120 U/ml when assayed by the present method for a2-PI activity. The activity was diminished to 55 U/ml by freezing and thawing the same sample of a2-M. However, a freezing-thawing procedure did not affect appreciably the inhibitor activity of purified a2-PI. Therefore, freezing and thawing the test sample before the assay decreases the activity by approximately 5-10lo and makes the assay method more specific to a2-PI. The decrease of inhibitory activity of a2-M to proteolytic enzymes (trypsin and chymotrypsin) by freezing and thawing procedures has been reported also by others (16). Since the presence of plasminogen in the test sample influences the assay method and gives erroneously lower values of inhibitor activity, plasminogen in the test sample was removed before the assay by lysine-Sepharose. 1 ml of the test plasma or serum, once frozen and thawed, was applied on a column of lysineSepharose (1 ml) equilibrated with barbital-buffered saline. The plasma was allowed to go into the column slowly and was followed with barbital-buffered saline. The first 5 ml of the effluent, which contained more than 95% of the protein applied, was collected. This fivefold-diluted sample was used for the assay. An aliquot of the sample was diluted in the test tube (10 x 75 mm) to 0.7 ml with cold barbitalbuffered saline and was mixed with 0.1 ml of 2% Cohn's fraction I of human plasma which contained 1.6 g/100 ml of fibrinogen and 1.2 casein U/ml of plasminogen. The mixture. was cooled in an ice-water bath. To this mixture were added 0.1 ml of 600 CTA U/ml urokinase (Mochida Pharmaceutical Co., Tokyo), and 0.1 ml of 20 U/ml bovine thrombin to make a final clot of 1 ml. The tube was quickly shaken to mix the reagents well and placed in a 37°C water bath. A stopwatch was started at the time of addition of thrombin. Soon there formed a clot in which many small air bubbles were fonned and trapped. The clot lysis was made noticeable by the sudden rise of the air bubbles to the upper surfaces. In another set of experiments, similar clot lysis tests without test sample were performed with various concentrations of urokinase. A standard curve was constructed by plotting on double-logarithmic paper the clot lysis time obtained by various concentrations of urokinase against units of urokinase in a 1-ml clot. A straight line resulted over the range of 6-60 U urokinase/ml clot. Using this standard curve, clot lysis time obtained with a sample containing the inhibitor was converted to a value of urokinase activity. There was an inverse relationship between the amount of the inhibitor present and resulting residual urokinase activity, which is linear in the region from 60 urokinase U (control with urokinase and no inhibitor) to 20 U/ml clot (urokinase plus various amounts of inhibitor). The inhibitor units per milliliter plasma or serum were calculated as the difference between the measurement of urokinase activity in the presence and in the absence of the test sample, expressed as units per milliliter clot divided by the volume (milliliters) of the test sample introduced into the assay system and multiplied by 5 (the dilution factor derived from the initial treatment of plasma or serum with lysine-Sepharose). The mean ±SD of the activities of a2-PI in plasma from 29 healthy Japanese adults measured by this method were 1,224+374 U/ml. Specificity of the clot lysis method for a2-PI. The IgG fraction of rabbit antiserum to a2-PI or a2-M was added to lysine- Sepharose-treated and frozen-thawed plasma, incubated at 37°C for 1 h and further at 4°C for 24 h. The resulting precipitates were removed by centrifugation. The residual inhibitor activity in the supernate was measured by the clot lysis method. 2, 1, 0.5, and 0.25 mg of IgG fraction of anti-a2-PI antiserum neutralized 95, 90, 82, and 70% of the inhibitor activity of 1 ml plasma, respectively. The similar amounts of IgG fraction of anti-a2-M antiserum did not appreciably affect the inhibitor activity. This indicates that the assay method is specific to some extent for a2-PI, but the effect of any other proteinase inhibitor in plasma on the assay can not be excluded. Thrombin preparation used in the assay was crude material and might have been loaded with nonthrombin biologic activities. However, when purified thrombin prepared by the method of Lundblad (17) was used instead of crude thrombin, it gave the same results as those obtained with crude thrombin, suggesting that any contaminant in the crude thrombin preparation did not affect the assay. Radioiodination of plasminogen. Plasminogen was radioiodinated by modification of the method of Greenwood et al. (18). To 13 mg of plasminogen in 6 ml of Tris-saline (0.05 M Tris-HCl, pH 7.4, 0.15 M NaCl) were added 0.15 mg of Nal in 0.15 ml of 0.05 M phosphate buffer (pH 7.4) and 80 ,uCi of 125I-Na solution (The Radiochemical Centre, Amersham, England) while stirring in an ice water bath. Immediately thereafter, 0.4 mg of chloramine-T (Tokyo Kasei Co., Ltd., Tokyo) in 0.1 ml of phosphate buffer was added, and the mixture was stirred for 1 min. Then 0.1 ml of Na2S202 solution (4 mg/ml phosphate buffer) was added to stop the reaction. Separation of 1251-plasminogen from the reaction mixture was carried out by gel filtration through a column of Sephadex G-50. Elution was initially carried out with 0.6 ml of the mixture of NaI (0.4 mg) and KI (2 mg), followed by. Behavior of arPlasmin Inhibitor in Fibrinolytic States. 363.
(5) Tris-saline. Specific activity of plasminogen before and after component disappeared after the plasma was absorbed radioiodination were 23.1 and 20.2 U/mg protein, re- with antiplasminogen antiserum, indicating that it is spectively. Distribution of '251-plasminogen or plasmin in urokinase- a complex between a2-PI and plasmin generated by activated plasma. 20 ul of radioiodinated plasminogen urokinase. When antigen-antibody-crossed electro(0.91 mg/ml) was added to 2 ml plasma, yielding plasma phoresis was carried out using antiplasminogen anticontaining 9 ,ug radioiodinated plasminogen/ml. This plasma serum, the appearance of another new component was activated with various amounts of urokinase and was with a faster mobility was found in the plasma subjected to electrophoresis in agarose in the same way as the first electrophoresis of antigen-antibody-crossed electro- activated with high concentrations of urokinase (> 100 phoresis. Immediately after the completion of electro- U/ml) as indicated by arrow 2 in Fig. 2B. This phoresis, the gel was cut transversely into 3-mm slices. component became less prominent when the plasma Each slice was put into a counting vial, and radioactivity was absorbed before electrophoresis with anti-a2-M was counted by Auto Well Gamma System, Aloka JDC-752 antiserum, indicating that the component is the com(Aloka Co., Tokyo). Statistical analysis. Statistical analysis was performed plex of a2-M and plasmin. It is noteworthy that the by Student's t analysis for paired or unpaired samples where complex of a2-M and plasmin did not develop when applicable. For unpaired samples, Snedecor and Cochran's the concentration of urokinase was as low as 25 U/ml modification was applied (19). A P value > 0.05 was con- (Fig. 2B), whereas the formation of the complex of sidered to represent a statistically nonsignificant change. The a2-PI and plasmin was observed (Fig. 2A). two-tailed test was used unless otherwise indicated. Patients. Effects of intravenous infusion of urokinase on plasma plasmin inhibitors were studied on seven patients who received urokinase therapy within 72 h after an attack of cerebral infarction. 240,000 U of urokinase was administered to each patient except for one patient who received 120,000 U. Patients having disseminated intravascular coagulation (DIC) with fibrinolysis were studied for the concentrations of plasmin inhibitors. The diagnosis of DIC was based on clinical observations as well as laboratory findings such as increased fibrin/fibrinogen degradation products (16-640,ug/ ml, normal < 4 ,ug/ml ), decreased plasma antithrombin III activity (10-80% of normal standard, normal range 90110% [21 ]), prolonged prothrombin time (> 15 s,.normal 1213 s ), decreased plasminogen (0.2-1.4 casein U/ml, normal range 1.4-2.0 casein U/ml), decreased fibrinogen gmmm RR', (20-200 mg/100 ml, normal range 150-400 mg/100 ml), and decreased platelet count (< 150,000/,A, normal range 150,000400,000/,ul) measured by Coulter counter (Coulter Electronics Inc., Hialeah, Fla.). These values were obtained by the repeated assays during the course of the diseases rather than by a determination on a single occasion. The gradual or sudden decline of platelet count, fibrinogen, plasminogen, and antithrombin III activity, together with further prolongation of prothrombin time and persistent presence or a steady increase of fibrin/fibrinogen degradation products, was most suggestive of DIC. Underlying diseases were carcinoma, acute leukemia, and septicemia.. l. :.. un. RESULTS. Development of the complexes of plasmin and inhibitors in urokinase-activated plasma in vitro. Plasma was incubated with various amounts of uro- FIGURE 2 Formation of the complex of the plasma inhibikinase, and subsequently subjected to antigen-anti- tors and plasmin in plasma activated with urokinase in body-crossed electrophoresis. Control plasma without vitro, demonstrated by antigen-antibody-crossed electroaddition of urokinase contained a2-PI as a single phoresis. To each 1 ml of plasma was added 2.5 and 50 component (Fig. 2A). When plasma was activated pA of urokinase solution (10,000 U/ml) separately, and the were incubated at 37°C for 2 h and further at room with urokinase, there appeared a new component mixtures temperature for an additional 12 h. They were subsequently which possessed the antigenicity of a2-PI with a slower subjected to antigen-antibody crossed electrophoresis using mobility in the electrophoretic field as indicated by antisera against a2-PI (A) or plasminogen (B). The direction arrow 1 (Fig. 2A). The new component became more of the first electrophoresis was from the left to the right. The (units) of urokinase added to 1 ml plasma are indiprominent, and the a2-PI component itself was dimin- amounts cated between the panels. Arrows indicate the complex of ished when the amount of urokinase was increased, plasmin and inhibitors. Arrow 1, the complex of a2-PI as seen in the bottom figure of Fig. 2A. The new plasmin. Arrow 2, the complex of a2-M and plasmin. and 364 N. Aoki, M. Moroi, M. Matsuda, and K. Tachiya.
(6) When these urokinase-activated plasma samples were analyzed by the crossed electrophoresis using antiserum against a,-AT, no formation of the complex of a1-AT and plasmin was observed, even in plasma activated with high concentrations of urokinase, such as 500 U/ml plasma. In these plasma samples, nearly all plasminogen was converted to plasmin, and most of the a2-PI formed complexes with plasmin (Fig. 2). The formation of the complex of a,-AT and plasmin was observed only after the addition of exogenous plasmin to this highly activated plasma. Plasma containing radioiodinated plasminogen was activated with urokinase, and the distribution of radioactivity in plasma fractions was analyzed by electrophoresis. The results are shown in Fig. 3. When plasma was activated with 25 U of urokinase/ml plasma, the peak of radioactivity shifted to the location (slice 5) corresponding to the peak of a2-PI-plasmin complex indicated by arrow 1 in Fig. 2A (Fig. 3A). The peak of a2-PI-plasmin complex became prominent with concurrent decrease of the original plasminogen peak (slice 3) when amounts of urokinase used were increased (Fig. 3B). With increasing urokinase, there developed another new peak of radioactivity (slice 9) which corresponds to the location of the peak indi-. E(a.. a. -11.. .7. >o[ <. IOO. -NoNDEPLETION X2M DEPLETED 2. 4 6 8 10 12 14 SLICE NUMBERS. FIGURE 4 Depletion of plasmin-a2-M complex by antiserum against a2-M. 0.1 ml of antiserum was added to 1 ml plasma which contained radioiodinated plasminogen and had been incubated with 125 U/ml urokinase as in Fig. 3. The mixture was then incubated at 37°C for 30 min and further at 40C for additional 12 h. The resulting precipitates were removed by centrifugation and the supernate was subjected to electrophoresis. The condition of the electrophoresis and the subsequent treatment were the same as in Fig. 3. The control (nondepletion) was run by replacing rabbit antiserum with unimmunized rabbit serum.. cated by arrow 2 in Fig. 2B (Fig. 3B). This new peak is most likely attributable to the formation of the a2-M-plasmin conmplex, in that the peak was destroyed by the treatment of the plasma with anti-a2-M antiserum before electrophoresis (Fig. 4). Development of the complexes of plasmin and inhibitors in plasma of patients receiving urokinase intravenously. For therapeutic purposes 240,000 or 300 120,000 U of urokinase together with 5,000 U of was infused intravenously for 2 h to seven heparin UK 25UNITS X200 different patients with cerebral infarction. ImmediA t ately before the infusion, and at 0, 3, 6 and 18 h after the termination of the infusion, blood samples 100 were withdrawn from the patients and subjected to analyses by antigen-antibody-crossed electrophoresis 2 4 6 8 10 12 14 (Fig. 5). The complex of a2-PI and plasmin was obSLICE NUMBERS served prominently in the samples withdrawn immedi30 ately after the infusion (Fig. 5). The complex became less prominent at 3 h, and no complex was found at 6 h after the infuision (Fig. 5). The antigen concentration, as well as the activity of a2-PI, decreased concurB rently with the appearance of the complex of a2-PI p 100oK 125 UNITS and plasmin. The decrease of activity was more remarkable than that of antigen concentration at the time nIU 2 4 6 8 10 12 14 immediately after the infusion (Fig. 5). The complex SLICE NUMBERS of plasmin and a2-M or a,-AT as described in in FIGuRE 3 Distribution of radioactivity in plasma fractions vitro studies mentioned above was not found at any after activation of plasma containing radioiodinated plas- time. minogen with urokinase. To each 1 ml of plasma was added Decrease of a2-PI after the intravenous infusion of 2.5, 12.5, and 50 ,ul of urokinase solution (10,000 U/ml) urokinase. To continue our studies in these patients, separately, and the mixtures were incubated at 37°C for 2 h concentrations (antigen) of the inhibitors, a2-PI, and further at room temperature for additional 12 h. They were subsequently subjected to electrophoresis in agarose, a2-M, and a,-AT, in plasma were measured immunoand the distribution of radioactivity was analyzed. The chemically before and after the intravenous infusion amounts of urokinase added to 1 ml plasma are indicated on (2 h) of 240,000 U of urokinase together with 5,000 U each figure. Control was run without addition of urokinase. of heparin. Inhibitory activity of a2-PI, a2-M antiSlices of agarose plate are numbered from the origins where the samples were applied. Details of the method are plasmin activity, plasminogen, and fibrinogen were also assayed. The concentrations and the activity of described in the text. UK, urokinase. 365 Behavior of a2-Plasmin Inhibitor in Fibrinolytic States CONTROL. u. r. 0. In. .. 1.
(7) fi.Af.41. FIGuRE 5 Formation and disappearance of the complex of a2-PI and plasmin in plasma of the patients under urokinase treatment, demonstrated by antigen-antibody-crossed electrophoresis using antiserum against a2-PI. Plasma samples were withdrawn from the patients before (a), immediately after (b), and 6 h after (c) the infusion of urokinase. Patients A and B received 120,000 and 240,000 U of urokinase, respectively. Arrows indicate the complex of a2-PI and plasmin. The concentration (mg antigen/100 ml) and the activity (U/ml shown in parentheses) of a2-PI are indicated under each figure.. a2-PI were significantly decreased by the infusion of urokinase and had not yet returned to the original values when examined at 18 h after the termination of infusion (Tables I and II). On the contrary, the concentrations of a2-M and a1-AT were not changed significantly by the infusion. Some significant increase of a2-M antiplasmin activity was noticed at the time immediately after the infusion (Table II). Plasminogen TABLE I Changes in the Concentrations of Inhibitors in Plasma after a 2-h Urokinase Infusion. decreased significantly after urokinase infusion, whereas fibrinogen remained unchanged (Table III). Decrease of atrPI in DIC with fibrinolysis. Concentrations of a2-PI, a2-M, and a,-AT in plasma were measured immunochemically in patients having DIC with fibrinolysis, and compared with normal values. The mean +2 SD of the concentrations of a2-PI in plasma from 25 healthy Japanese adults were 5.98 ±+1.76 mg/100 ml. The reported values of the mean ±2 SD of concentrations of a2-M and a,-AT in plasma TABLE II Changes in Inhibitor Activities of Plasma after a 2-h Urokinase Infusion. Time after the start of infusion, h. Inhibitors. 0. 2. 20. Time after the start of infusion, h. mg/100 ml. a2-PI. 4.78±+1.27*. a2-M. 191+22. a,-AT. 244+40. 3.42+0.52 P<0.02t. 3.02±0.29. 191+19 NS 248+41 NS. 198+26 NS 261+24 NS. 0. P<0.02. * Mean+SD (n.= 6). t P values refer to the difference in concentrations in samples at 2 or 20 h from those in samples at 0 time.. 366. Inhibitors. 2. 20. U/mi. a2-PI a2-M antiplasmins. 1,461+298*. 588±206 P<0.001t. 103+22. 136± 18 P < 0.05. 911+244 P<0.005 95+22 NS. * Mean+SD (a2-PI, n = 6; a2-M, n = 5). t P values refer to the difference in activities in samples at 2 or 20 h from those in samples at 0 time.. N. Aoki, M. Moroi, M. Matsuda, and K. Tachiya.
(8) TABLE III Changes in the Concentrations of Plasminogen and Fibrinogen in Plasma after a 2-h Urokinase Infusion Time after the start of infusion, h 0. 2. 20. Plasminogen, casein U/ml. 1.41±0.26*. 1.06±0.22 P<0.0005t. 1.05±0.3 P<0.05. Fibrinogen, mg/lOO ml. 394±123. 337±91 NS. 346±86 NS. * Mean±SD (n = 5). t P values refer to the difference in concentrations in samples at 2 or 20 h from those in samples at 0 time. One-tailed test.. from 160 healthy Japanese adults measured by single radial immunodiffusion are 210+100 and 200+100 mg/100 ml, respectively (23). The concentrations of a2-PI in most of the patients were found to be lower than the normal range, whereas values of a2-M were in the normal range except for one case, and those of a,-AT were higher in many cases than the normal range (Fig. 6). Many of the values of a2-M, even though in the normal range, were well below the mean level of normals, thus implying that there had been some reduction in the level in those cases. No complex of plasmin with any one of the inhibitors was detected when these patients' plasmas were analyzed by antigen-antibody-crossed electrophoresis. a2 -PI. Ot2 -M. 8r. 400o.. 800. 6. 300 -. 600. 0. 0. *0 0. 0. *.0. -40-. E. 4001. 200-. C!4. *t. 1001-. 2. ---. I a___. 2001-. 0. OL. O1 P< 0.001. o0 NS. FIGURE 6 Concentrations of a2-PI, a2-M, and al-AT in plasma of the patients with DIC, measured by the immunochemical method. The mean concentration of each inhibitor is indicated by a short horizontal line. The means +2 SD of normal values for healthy adults are indicated by solid and interrupted horizontal lines, respectively. P values refer to the difference in concentrations in DIC from those in normal.. DISCUSSION. The concentration ofa2-PI in normal plasma was found to be approximately 5-7 mg/100 ml, which is calculated to be _0.7-1.0 ,uM on the basis of mol wt 67,000. The molar concentrations of a2-M (mol wt 820,000) and al-AT (mol wt 54,000) in normal plasma are estimated on the average at 2.4 and 37 ,uM, respectively, while their average concentrations are 210 and 200 mg/100 ml, respectively (23). Consequently, it may be concluded on the assumption of one-to-one molar reaction of inhibitor and enzyme that a2-PI contributes only 2.5% on average of the total capacity of antiplasmin activity of the whole plasma. However, a2-PI has the strongest affinity for plasmin among these inhibitors. It had been suggested that the strong affinity of a2-PI for plasmin plays an important role in the regulation of fibrinolysis. To test this possibility, the behavior of a2-PI in fibrinolytic states was explored in the present study. When plasminogen in plasma was activated in vitro by a relatively low concentration of urokinase, there appeared readily a plasmin-a2-PI complex. However, no complex of plasmin-a2-M or a,-AT was observed. The formation of the plasmin-a2-M complex was observed only when plasma was activated with a larger amount of urokinase and after most of the a2-PI was consumed by forming complexes with plasmin. The formation of plasmin-al-AT complex was not observed even when all of the plasminogen in plasma was converted to plasmin unless exogenous plasmin was added. When a relatively low dose of urokinase was infused intravenously to patients, the consistent findings were the decrease of the concentration and the activity of a2-PI and the concomitant formation of the plasmina2-PI complex. The decrease of the activity of a2-PI was more remarkable than that of the concentration at the time of the existence of the complex in blood. This can be explained by the fact that the concentration measured by the immunochemical method includes the enzyme-inhibitor complex which has lost the activity. In contrast to a2-PI, concentrations of a2-M and a,-AT were not significantly changed, and no complex of plasmin-a2-M or -a,-AT was observed. A significant increase of a2-M antiplasmin activity observed immediately after the infusion might be attributable to antiplasmin activity of heparin-antithrombin III (24), because there was no significant increase of a2-M antigen concentration throughout the study, and no significant increase of a2-M antiplasmin activity was observed 18 h after the termination of infusion when no residual effect of heparin was detected. Arnesen and Fagerhol studied effects of urokinase in-. Behavior of a2-Plasmin Inhibitor in Fibrinolytic States. 367.
(9) fusion on a2-M, a,-AT, and antithrombin III in plasma (25). The doses of urokinase that they used were far greater than those used in the present study. They infused nearly twofold the amount of urokinase as compared to ours in 10 min as a loading dose and continued infusion of a maintenance dose for 18h. Their hourly maintenance dose is nearly equivalent to the total amount of urokinase infused into each patient in the present study. In spite of their large dose, the decrease of a2-M after completion of infusion was only 20% on average of the preinfusion values. There was no change of plasma antithrombin III level, and the concentration of a,-AT was even increased. Nilehn and Ganrot activated practically all plasminogen by infusing a large dose of streptokinase and observed the decrease of a2-M to 50% of the initial values (26). In the present study, the decrease of plasminogen level averaged 25% of the preinfusion values, and no decrease of a2-M was observed. These data, together with those obtained by others (27, 28), indicate that a2-M is one of the major antiplasmins which react with plasmin generated during fibrinolytic therapy. The present study provides, furthermore, information on a newly discovered inhibitor in addition to a2-M. The present study suggests that a2-PI is the most reactive plasmin inhibitor in plasma, and whenever plasminogen activation takes place, a2-PI is the first inhibitor which reacts with plasmin formed. Only when the excess amount of plasmin is formed, and when most a2-PI has been consumed by forming complexes with plasmin, then a2-M will become the principal inhibitor to plasmin. This view is supported by the finding that, among the major plasmin inhibitors in plasma, a2-PI was the only inhibitor of which the concentration was significantly decreased in patients with DIC. However, no complex of plasmin-a2-PI was observed in plasma of these patients, at least by the method presently employed. This is explained by the rapid disappearance of plasmin-a2-PI complex from the circulating blood, which, in the present study, was indeed shown after the infusion of urokinase. The plasmin-a2-PI complex formed after intravenous infusion of 240,000 U of urokinase disappeared from the circulating blood within 6 h. The complex was probably removed by the clearance mechanism of the reticuloendothelial system in organs especially in the liver, since the complex is formed by a stable covalent bond (4), and its dissociation can not readily occur. a2-M might also have reacted with plasmin in these DIC patients, inasmuch as almost all values of a2-M were below the mean value although they were in the normal range. However, this possibility was not statistically proved in the present study (P > 0.1). The cause of the increase of a,-AT observed in these DIC 368. patients was not known, but it is interesting to note that a,-AT concentration in plasma was steadily increasing during prolonged infusion of urokinase (25). Because the molecular weight of the complex of a2-PI and the light chain of plasmin was previously reported as 84,000 (4) and the molecular weight of the heavy chain was nearly 67,000 (4), the molecular weight of plasmin-a2-PI complex must be close to 150,000. The plasmin-a2-PI complex might be the same as the complex reported by Collen et al. (29). They reported that, during the activation of plasminogen in plasma with urokinase in vitro, there appeared the complex of plasmin and a hitherto unidentified antiplasmin. The molecular weight of this complex was estimated by gel filtration to be in the region of 130,000-170,000, which is the same region of the estimate of molecular weight of plasmin-a2-PI complex. Miillertz (30) also reported that urokinaseactivated plasma contained, at a low concentration of urokinase, a component with plasmin + plasminogen antigenicity which was eluted with nearly 7S protein from Sephadex G-200 and had 8-mobility by gel electrophoresis. This component appears from its physical properties to be identical with plasmin-a2-PI. complex. In addition to these observations, plasma inhibitors of fibrinolysis which possess properties similar to those of a2-PI were recently described (31-35). The question as to whether or not a2-PI is identical with any one of these inhibitors needs further investigation, including elucidation of the physicochemical and immunochemical properties of each inhibitor. Crossed immunoelectrophoresis of some plasma samples using anti-a2-PI antisera gave double immunoprecipitate lines as seen in patient B in Fig. 5. Development of double immunoprecipitate lines might be due to the molecular heterogeneity of antigen as suggested by Laurell (36), and the antisera might contain antibodies against more than one of the determinants. Molecular heterogeneities of a2-M and al-AT have already been revealed by crossed immunoelectrophoresis (37). Another possibility that the antiserum is not specific and reacts with two different molecules which possess the same electrophoretic mobility cannot be excluded, although immunoelectrophoresis and double immunodiffusion have not been able to demonstrate this possibility (4).. ACKNOWLE DGMENTS The authors would like to thank Dr. M. Yoshida and Dr. A. Ueki of the Department of Neurology at the Jichi Medical School for their generous cooperation. Acknowledgment is also due to Dr. W. H. Seegers, Wayne State University, for his critical reading. This work was supported by a grant from the Ministry of Education of the Govemment of Japan.. N. Aoki, M. Moroi, M. Matsuda, and K. Tachiya.
(10) 1. 2. 3.. 4.. 5.. 6. 7.. 8. 9.. 10. 11. 12. 13.. 14. 15.. 16.. 17.. 18.. 19. Snedecor, G. W., and W. G. Cochran. 1967. Statistical REFERENCES Methods. Iowa State University Press, Ames, Iowa. 6th Aoki, N., and K. N. von Kaulla. 1971. Human serum edition. 114-116. plasminogen antiactivator: its distinction from anti- 20. Merskey, C., G. J. Kleiner, and A. J. Johnson. 1966. plasmin. Am. J. Physiol. 220: 1137-1145. Quantitative estimation of split products of fibrinogen Aoki, N., and T. Kawano. 1972. Inhibition of plasminogen in human serum, relation to diagnosis and treatment. activators by naturally occurring inhibitors in man. Am. J. Blood. 28: 1-18. Physiol. 223: 1334-1337. 21. Monkhouse, F. C. 1963. The influence of inorganic salts Aoki, N., and M. Moroi. 1974. Distinction of serum inon plasma antithrombin activity. Thromb. Diath. hibitor of activator-induced clot lysis from a,-antiHaemorrh. 9: 387-394. trypsin. Proc. Soc. Exp. Btol. Med. 146: 567-570. 22. Tocantins, L. M. 1964. Estimation of prothrombin (OneMoroi, M., and N. Aoki. 1976. Isolation and characterizastage method of Quick). In Blood Coagulation, Hemortion of a2-plasmin inhibitor from human plasma. A novel rhage and Thrombosis. L. M. Tocantins and L. A. Kazal, proteinase inhibitor which inhibits activator-induced clot editors. Grune & Stratton, Inc., New York. 148-150. lysis. J. Biol. Chem. 251: 5956-5965. 23. Matsuda, T. 1967. Coagulation studies on strokes. Nihon Brockway, W. J., and F. J. Castellino. 1972. MeasureRinsho. 34: 72-78. ment of the binding of antifibrinolytic amino acids 24. Highsmith, R. F., and R. D. Rosenberg. 1974. The inhibito various plasminogens. Arch. Biochem. Biophys. 151: tion of human plasmin by human antithrombin-heparin 194-199. cofactor.J. Biol. Chem. 249: 4335-4338. Robbins, K. C., and L. Summaria. 1970. Human plasmin- 25. Arnesen, H., and M. K. Fagerhol. 1972. c2-Macroglobulin, ogen and plasmin. Methods Enzymol. 19: 184-199. a1-antitrypsin, and antithrombin III in plasma and serum Barlow, G. H., L. Summaria, and K. C. Robbins. 1969. during fibrinolytic therapy with urokinase. Scand. J. Molecular weight studies on human plasminogen and Clin. Lab. Invest. 29: 259-263. plasmin at the microgram level. J. Biol. Chem. 244: 26. Nilehn, J-E., and P. 0. Ganrot. 1967. Plasmin, plasmin 1138-1141. inhibitors and degradation products of fibrinogen in huAlkjaersig, N., A. P. Fletcher, and S. Sherry. 1959. The man serum during and after intravenous infusion of mechanism of clot dissolution by plasmin. J. Clin. streptokinase. Scand. J. Clin. Lab. Invest. 20: 113-121. Invest. 38: 1086-1095. 27. Fischer, M. 1971. Influence of thrombolytic therapy on Remmert, L. F., and P. P. Cohen. 1949. Partial purihuman antiplasmins. Thromb. Diath. Haemorrh. 47 fication and properties of a proteolytic enzyme of human (Suppl.): 153-157. serum.J. Biol. Chem. 181: 431-448. 28. Spottl, F., and F. Holzknecht. 1970. The influence of Ratnoff, 0. D., and C. Menzie. 1951. A new method for inhibitors of plasmin and plasminogen activation on the the determination of fibrinogen in small samples of streptokinase-induced fibrinolytic state. Thromb. Diath. plasma. J. Lab. Clin. Med. 37: 316-320. Haemorrh. 24: 100-112. Rowe, D. S., and J. L. Fahey. 1965. A new class of human 29. Collen, D., F. De Cock, and M. Verstraete. 1975. immunoglobulins. II. Normal serum IgD. J. Exp. Med. Immunochemical distinction between antiplasmin and 121: 185-199. alpha1-antitrypsin. Thromb. Res. 7: 245-249. Laurell, C. B. 1966. Quantitative estimation of proteins 30. Mullertz, S. 1974. Different molecular forms of plasminby electrophoresis in agarose gel containing antibodies. ogen and plasmin produced by urokinase in human Anal. Biochem. 15: 45-52. plasma and their relation to protease inhibitors and Laurell, C. B. 1965. Antigen-antibody crossed electrolysis of fibrinogen and fibrin. Biochem. J. 143: 273-283. phoresis. Anal. Biochem. 10: 358-361. 31. Hedner, U. 1973. Studies on an inhibitor of plasminogen Matsuda, M., S. Iwanaga, and S. Nakamura. 1972. A activation in human serum. Thromb. Diath. Haemorrh. simple, large scale method for preparation of plasmino30: 414-424. gen-free fibrinogen. Thromb. Res. 1: 619-630. 32. Mui, P. T. K., H. L. James, and P. Ganguly. 1975. IsolaIwamoto, M., and Y. Abiko. 1970. Plasminogen-plasmin tion and properties of a low molecular weight antisystem. Preparation of a2-macroglobulin antiplasmin plasmin of human blood platelets and serum. Br. J. from human plasma. Biochim. Biophys. Acta. 214: Haematol. 29: 627-637. 402-410. Gallimore, M. J. 1975. Serum inhibitors of fibrinolysis. Lebreton de Vonne, T., H. Mouray, G. Berthillier, and 33. Br. J. Haematol. 31: 217-231. R. Got. 1972. Influence du vieillissement des a2-macro- 34. Hedner, U., and D. Collen. 1976. Immunochemical disglobulines de lapin sur la formation du complexe a2tinction between the inhibitors of plasminogen activamacroglobuline-enzymes. Biochim. Biophys. Acta. 257: tion and antiplasmin in human plasma. Thromb. Res. 365-371. 8: 875-879. Lundblad, R. L. 1971. A rapid method for the purification of bovine thrombin and the inhibition of the puri- 35. Edy, J., F. De Cock, and D. Collen. 1976. Inhibition of plasmin by normal and antiplasmin-depleted human fied enzyme with phenylmethylsulfonyl fluoride. Bioplasma. Thromb. Res. 8: 513-518. chemistry. 10: 2501-2506. Greenwood, F. C., W. M. Hunter, and J. S. Glover. 36. Laurell, C. B. 1972. Electroimmunoassay. Scand. J. Clin. Lab. Invest. 29 (Suppl. 124): 21-37. 1963. The preparation of 1311-labelled human growth hormone of high specific radioactivity. Biochem. J. 89: 37. Ganrot, P. 0. 1972. Crossed immunoelectrophoresis. Scand. J. Clin. Lab. Invest. 29 (Suppl. 124): 39-47. 114- 123.. Behavior of a2rPlasmin Inhibitor in Fibrinolytic States. 369.